The principle of charity: Criticize the best interpretation of an argument
effectiviology.com
effectiviology.com
For those who've not read the article and are thinking "but what about <insert political scumbag here>", the article starts by saying you should be charitable "as long as it’s reasonable" to do so. It is reasonable to assume that a political operative of almost any persuasion will attempt to bend things and miss-represent, so you should be on the look out for that and treat it accordingly.
I like the idea of the "steel man" argument, as a more effective alternative to the "straw man", you end up with a better argument that way, and are far more likely to "win" the argument than if you present something that the other person sees as BS. Even if your opponent is talking BS, the counter-argument you come up with will be more effective in countering it, a weak argument predicated on the weakest interpretation gives them wriggle-room.
I feel like the whole idea is the debate equivalent of the idea "never attribute to malice that which can be explained by incompetance". Sometimes it is malice, but usually it's just a mistake. Assuming the worst in people as the default is very isolating, and understanding that made a big difference to my state of mind as I got older.
That's actually a Russian proverb, afaik. In German the equivalent typically is attributed to Lenin ("trust is good but supervision is better"), but I think without source.
Generally I found that if you explain to someone how their comment might have been out of place, or maybe they just did not consider a nicer interpretation you get a nice response back.
> "you are an X, therefore your argument is junk because you are stupid/evil, here is why"
The problem with this type of comments is that generally people are not very good at complaining about them, or at suggesting better way to frame criticism. This is especially relevant in internet forums where too many personalities interact to be able to generalize and it is hard to gauge how much criticism is warranted.
I do feel like it's going to be realy important to teach my kids as they grow up, how to have good conversations online,and how these sort of mediums often twist the perseption of what they say without them realising.
Reputation matters. Filtering matters. Selectivity matters.
We are always reasonable in our judgement.
Of course, this also means that a lot of "unacceptable in normal discourse but supportable via evidence" positions get pushed. Which quite a few people seem to find disagreeable. (you see, only MY outside-the-overton-window ideas are worthy of discussion - everything else is horrible and you're a Nazi/Commie/Racist/Sexist/etc for even listening to it or allowing it to be discussed by others. How dare you!)
A student is not bound to reproduce slavishly the interpretations of others, but you are bound to assess critically what you read.
Before you can say, "I disagree," you must be able to say, "I understand." It is axiomatic that before you level criticism you should be able to state an author's position in terms that he or she would find acceptable. After that, you may proceed in any of six directions:
a. Show where the author is misinformed.
b. Show where the author is uninformed.
c. Show where the author is inconsistent.
d. Show where the author's treatment is incomplete.
e. Show where the author misinterprets through faulty assumptions or procedures.
f. Show where the author makes valuable contributions to the discussion at hand.
[1] Gordon Fee, Exegesis, 3rd edition 2002 p.33
This assumes that the assumption of good faith is well founded. Life is too short to critically refute every single crackpot theory out there. For better or worse, if someone sufficiently damages their ethos, I will choose to not spend my time critically evaluating everything they say.
[edit]
One place where my argument does not apply is when large numbers of people hold a particular opinion. If e.g. 80% of Republicans or 80% of Democrats hold a view, we are talking about 1/3 of the population of the US. Understanding why they hold that view, no matter how non-sensical it may seem is vital to understanding the world we live in.
It is insufficient to just say "well then insert large group here are all idiots" because that is clearly not true.I have two family members who each got perfect SAT scores and one is about as hippy-left as you can get and the other is a card-carrying republican and Trump supporter. They are (in my opinion) both wrong about a lot of things, but dismissing them as idiots is not possible.
In this case, rather than claiming you disagree (which should imply understanding), you can just say you don't know enough about the issue to comment on it.
The spirit of the above argument is that, if you're going to vehemently disagree or argue with someone, you should understand— not that you need to understand every issue, nor agree or disagree with every issue.
aidenn0 is making a subtlety different point: this form of rational argument depends on good-faith. If you think that has been violated, you would not say "I don't know enough about the issue to comment on it," but rather "I don't think this person is arguing in good-faith."
I agree with the principle of charity, and your claim that you should understand before disagreeing. But it is also true that you will encounter people who are not arguing in good-faith, in which case it is sometimes a mistake to even engage.
On the internet (which is where the quality of rhetoric could most use the principle of charity these days), it is hard or even impossible to tell the difference between a troll or a crackpot.
The issue is that these rules can be exploited by a bad faith actor. Just throw arguments at somebody rapid fire, a gish gallop, and it basically works like a DOS attack. Eventually your opponent will stop responding after which you can declare victory since your opponent was unable to refute your argument. This incidentally is an argument against free speech absolutist positions.
I feel a lot of these rules assume you're debating an academic acting in good faith who wouldn't want to resort to such tactics for fear of hurting their own reputation. They're for people who argur with eachother by writing books about how their opponents are wrong. They're not adapted to the pseudoanonymous low barrier to entry environment of the internet.
You're best off taking shortcuts that result in you making weak arguments to avoid getting bogged down in discussions with bad faith/unintelligent actors and that allows you to be the among the first to opine on an issue which grants you a wider audience. At least to some extent.
https://news.ycombinator.com/newsguidelines.html
Before that, I used to write about the principle of charity quite a bit: https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme.... But eventually stopped using that name so much because it can be a bit confusing to people who haven't heard it before. It's shorter to just quote the guideline.
1: Take someone's position that one doesn't like
2: Extend it to a reasonable analogue (aka apply 1st formulation of Kant's Categorical Imperative)
3: Show how absurd it is and try to get the original person to defend it.
4: Never provide constructive criticism nor an alternative viewpoint.
It's a toxic way to entrap others and degrade the quality of a conversation. So much so that I had to write a satirical post about it[0]
You articulated something I've noticed before but never really expressed as concisely. If you have the time and inclination, you should write more longform pieces about this. I'd read them.
"1. Find an argument that is disagreeable."
Somebody does not agree that people should expect others to remember their preferred pronouns.
"2. Derive an underlying principle behind it. This must appear civil and reasonable at all costs as it gives you credibility. At all costs, do not engage with the actual argument."
This is the article itself, which is framing the discussion as an exposure on how people win a discussion by using a certain dark strategy (this is the really clever meta move!)
"3. Extend the principle to a plausible real world absurdity to demonstrate its impracticality."
A 7-point bullet-point list is provided, culminating on the suggestion that people who are pronoun-conservative possibly also want to commit genocide against gay people and Jews.
"4. Use this absurdity to: tear down the argument; or appeal to a slippery slope; or directly ask the original author to defend."
The subtle escalation along the bullet-point list is very clever, it slowly "boils the frog", distracting the reader from the slippery slope argument contained therein.
"5. At all costs, do not propose an alternate solution which would reveal the true intentions for doing this. That cannot be revealed."
Yup.
To illustrate what a more nuanced compromise might look like, I disclose my own position on this matter:
If somebody tells me that me addressing them with a given pronoun is important to them, I will make my best effort. At the same time, my brain has been primed by decades of following the convention of choosing pronouns based on perceived gender, so I also expect to be excused if I forget, and for the person to consider that I might be tired or have bad memory, instead of assuming malice on my part.
"6. If anyone else replies to counter or patch holes, repeat the steps at the top."
I guess I'm the one playing that role here. Your move. :)
My move is simple. Do nothing. My satire fails its own test when applied to itself as I make clear what I am looking for and what is unacceptable, emphasized with satire.
I clearly value an argument when folks make a good faith effort to argue for an alternative position, make constructive criticism, elaborate on the source of their concerns, or reveal their intentions. No hidden intentions from me.
As for the example: I thought the satire had made it very clear that the alternate solution is to have the would-be-saboteur reveal their alternate position, make constructive criticism, elaborate on the source of their concerns (ex: like you've done above), or reveal their intentions. So my satire also fails my own test that way. Again, no hidden intentions, clear alternatives.
Since you made your position clear, we can now have a constructive argument about a myriad of things: the role of perceived gender in society, cultural norms, morality of honest mistakes, etc. And I don't have to be guessing a position on your behalf while also defending a position I never took. If you let me pick for you, and you pick for me, that's not a discussion.
It's basically a giant post about intellectual honesty and good faith discussions.
There is certainly an argument to be made that producing this sort of document ought to be a (not "the", but "a") end-goal of more philosophical conversations. Probably created by multiple authors, though. I think there's a case to be made that while a one-sided screed will always have its place, there isn't enough work put into this sort of back & forth right in one place structure.
That said, the _Summa_ assumes that you are reasonably familiar with the works of Aristotle and Plato - so some arguments (or counter-arguments) may sound arbitrary to those who are not yet familiar with those works.
Final note, another good source to read the Summa is http://newadvent.org/summa/
It is essential to coherent conversation that you do your best to understand your interlocutor's position. If their argument betrays some particular idiosyncrasy or mistake, then you should focus on and examine that. If you instead ignore those features of their position and impose a view upon them which they don't actually possess - 'the best possible interpretation of their argument' - then you will not actually be engaging with their reasoning, and thus you'll be less likely to understand the source of your disagreement, analyse the merits of those differences, and convince one another.
You should not be looking for a 'good' or 'bad' interpretation of their argument, you should be searching for an accurate one.
I think better principles of charity would be:
1. Assume good faith 2. Be open minded to alternative views 3. Don't treat arguments as a battle of wits but as an attempt to move towards mutual understanding 4. Assume the person you're arguing with is at least as intelligent as you 5. Engage with their reasoning
The way to get the most value out of flawed people is to find the best version of their argument and then argue with that. The person isn't all that relevant.
I disagree. Human beings want to have their views recognized. Dismissing their views because there's a better position they could take is extremely presumptuous and doesn't generally make the conversation constructive. Because the other person will recognize their views aren't being heard, and won't necessarily care whether it's because of a misguided altruistic sense of bettering their argument for them.
I think this is completely true. Also, it fully validates the point you were responding to.
We can be in a situation where we simultaneously realize two things:
1/ the sole remaining obstacle to agreement is our discussion partner's ego, and
2/ our partner is not yet ready to confront the truth of 1/ for reasons outside of their control.
If we do realize these two things but continue to demand that the other side dissolve their ego right now or else they're not meeting our standard of good-faith (although 2/ clearly says they are), then that would be an example of our own ego getting in the way. Which shows that your criteria is valid and fully affirms the grandparent post.
'Simply put, the principle of charity is the idea that when criticizing someone’s argument, you should criticize the best possible interpretation of that argument.'
Later:
'Essentially, the principle of charity embodies the idea that when you interpret what other people say, you should select the best possible interpretation for their statements.'
What is at question in both these statements is the position of an interlocutor. It is a guide to conversation.
You say:
'The way to get the most value out of flawed people is to find the best version of their argument and then argue with that.'
That seems like a highly narcissistic way of conducting a conversation. You would in effect be having an argument with yourself.
As I said above, I think imposing a view on an interlocutor that they do not actually hold is a recipe for confusion, will almost certainly make the conversation unproductive, and might prevent you from understanding their position in the first place.
You give your argument, they give theirs. Then you amend their argument for them, and argue against that, ignoring your interlocutor.
If you are not interested in accurately understanding the person you're talking to, and instead impose what you take to be the strongest interpretation of their argument on them, then you are for all intents and purposes arguing not against them, but yourself.
The problem is that language is ambiguous and easily misunderstood. You interlocutor also has context which you don't know about.
Being charitable means recognizing these facts and so reading their words to mean the strongest possible interpretation. If they intended some weaker argument, it doesn't matter, because dismantling the stronger version already handles the weaker interpretation.
Take the following example. Suppose you claim that the meat industry is bad, advance three premises, and infer from them the conclusion that it is indeed bad. Now, if I use the dictum 'criticise the strongest version of your interlocutor's' argument, that could mean many things:
1. I might think that all the premises were unconvincing and argue against entirely different premises in favour of the same claim.
2. I might offer new premises in addition or substitution for the premises provided, or a different conclusion.
3. I might interpret the specific details of the premises or conclusion differently.
I personally would say that this conversation is based on an unhelpful conflation. We should both strive to understand one another as accurately as possible, and to move the conversation in as constructive a direction as possible. They need not be in tension. The problem arises if I take the latter - the desire to make the conversation constructive - as an interpretive principle.
The dictum applies prior to this step, where the premises and the argument may be phrased in a somewhat ambiguous way. When interpreting the language used to describe the premises, take the most challenging interpretation rather than the weakest/most easily attacked interpretation. When evaluation the inferences, once again take the most challenging interpretation.
For instance, a phrase may admit two interpretations, a valid or invalid inference. You should always assume your interlocutor intended the valid inference.
If you read carefully and allow the conversation to unfold, while prompting your interlocutor to explain what remains unclear to you, then generally you can get a fairly accurate impression of their views. More often than not, the problem is not ambiguity, but an unwillingness to listen, and engage thoughtfully with others, through a genuine dialogue.
> You should attempt to re-express your target’s position so clearly, vividly, and fairly that your target says, “Thanks, I wish I’d thought of putting it that way.”
It is not always the case, though, and you do yourself a disservice by assuming that it is.
It's not a static situation, and most of the outcomes are still better than if you drill down on some particular issue they have and play the "gotcha" game. (At least, better if your goal is authentic intellectual conversation.)
E.g. we know that social media are infected by a variety of trolls and "influencers", and it seems pointless to assume the best possible interpretation for their output - except, possibly, as a deliberate rhetorical position, which can sometimes be more persuasive than more passionate rebuttals.
Being a bad actor may chase someone out of an argument giving you an immediate “win,” but the word goes around about you, and suddenly your professional reputation is in tatters.
But in an environment where people are basically anonymous, where you may be engaging with a paid operative, or a throwaway sock puppet account, or an account that is trying to amass followers for grift (like being an anti-vaxxer online in order to sell books and newsletters)...
None of the traditional manners around debate apply.
The term for them is "useful idiots." But you definitely shouldn't engage with them like they are bad-faith actors, because often they're good-faith actors who've been mislead.
It's the commentary that isn't clear that it's paid, the 300 anonymous comments from different accounts on some news story all expressing some opinion that turn out to be from the same source, that's the real problem.
Whether on average you're dealing with paid operatives is hard to tell. Whether they exist is not; they definitely do. (Honestly, the alternative is just absurd. "There's billions of dollars worth of value sloshing around in Internet forums in setting the opinion and the flow of conversation, but everyone in the world is so very honest that the very idea of forging the signals simply never occurs to anyone, and nobody ever has a reason to try to slant things and is willing and able to pay." Come on. Seriously.) It's the scale that's at issue.
From personal experience I can testify that Israel was actually the first in the game, as I remember engaging against very persuasive pro-Israel posters on reddit during the summer of 2006 (meaning during Israel's war against Hezbollah), at a time when that site was almost totally politics-free. That was a lesson for me and I've since given up on any Israel-related discussions online.
The Russian shill accounts started showing up two summers later, in 2008, during Russia's war with Georgia. Somewhere in the reddit server logs there's all this information stored and ready to be analyzed but I suspect no-one has any real incentive in doing this.
While that may be true on the English web, I don't think that works as a general principle. There are definitely some contexts where it's much more likely than you estimate, e.g. https://gking.harvard.edu/50c.
Other times you end up looking like a soft-bellied fool.
This is by far the best rule I have ever attempted. The best way to debate is when both sides reword the other side argument, with corrections until both are satisfied.
It will also reveal the most insightful and depressing hidden realities about human cognition. Significant percentage of people are seemingly incapable of entertaining alternative viewpoints. Everyone suffers from this condition when we are heavily invested int some idea or belief, but most people can suppress the tendency temporarily and see the logic of some other way of thinking.
I wonder if there are psychological tests that can be used to score this ability.
In other words, the most common type of reply I get is something to the effect of, “Despite re-stating my argument with clear articulation, you did not understand it (and I won’t say why) and further, your counterpoints don’t constitute “evidence” or induce any information transfer (which I will take to imply your deficiency and not mine), and before looking further into any of this, the burden of collecting, organizing and disseminating evidence is all on you.”
This attitude seems often to get rewarded with upvotes as well. It’s very distressing to me personally.
The effect I mentioned can only be detected when you talk face to face with someone in good faith and you both try to come to conclusion.
This is also why people should challenge others in public when needed. Your boss might not change his ways, but your colleagues could.
In general I find those things to be very personally discouraging on Hacker News. It’s a community I’d like to belong to, but feel that when I abide very carefully by the community standard of discourse, I still receive punitive downvotes for disagreeing with crowdthink or receive canned moderator responses that don’t explain the specific mechanics by which a comment is considered failing to meet guidelines, with no serious appeals process when you believe you’re being treated unfairly.
Turns out that this is something lots of engineers could benefit from as this is simply not part of their bag of tools when they leave university and something that can frustrate them enormously when they need to engage with others whether it is colleagues, customers, or superiors.
I've been in more than a few meetings where people were yelling at each other, or worse, where I ended up doing a fair bit of yelling myself. These days, I tend to be a bit more conscious about it than I used to be. Usually it indicates something is wrong that needs fixing and quite often it is beyond whatever is being discussed. Fixing that is often the more important thing.
A few tricks I apply: 1) If you think something is good, make a point of communicating that. It buys you good will and people are more likely to listen to you when you say otherwise. Also it keeps people motivated. 2) Talk in terms of improvements/fixes/changes instead of simply stating, "this sucks". It's much harder to argue against that by others because why wouldn't you want to improve things. 3) Pick your fights carefully. Having opinions is cheap but owning the consequences is not. I often give people room to do things their way even when I don't fully agree it's the right way. Basically I need them to do it for me without my involvement. And there's always the chance you are wrong after all. In any case, the whole point of a team is being able to trust others to get shit done without you. If you need to micromanage everything, something is wrong. That something might be you.
Note well, however, that if I'm reading this, and I see that you have judged the other side as arguing in bad faith, and you try to document that, and I don't agree, then I think you have a hair trigger at least, and perhaps that you are arguing in bad faith. That is, you have to extend more charity than a normal reader, rather than less.
But most importantly i try to never point thd moral finger at someones position that way a much more fruitful discussion can be had.
There is a kind of corollary to this: Often times person A comes to me complaining about person B being stubborn about something. So I often ask: "So why do you think B is insisting on it?" They usually cannot come up with a good answer. At that point, I either hint or explicitly tell A that B has a position that A doesn't understand, and it's not surprising B will not change his position since A isn't trying to address B's needs.
It turns out even hardened criminals want to be heard/understood, and by using (author calls) "tactical empathy", you can start your negotiation off on the right footing.
I am usually trying to engage people in discussion, not "win" an argument. I try to add value to the discussion, not "win."
I especially dislike the general assumption that winning the argument means making the other person lose. Some people are seriously not happy with just being heard. No, they want someone else to go down in social rank or look like a fool. Without that element, they obviously aren't happy.
This has the virtue of not giving them the victory that they wanted. Even if they take the last word, it's usually pretty clear to an unbiased reader what was going on. I'm (at least sometimes) content to leave it there, and to let the reader judge.
Specially with pearls like this: https://scifi.stackexchange.com/questions/91393/dune-explain...
I like this and I'd even extend this to having an opinion on something. It really helps your opinions!
>> “I never allow myself to have an opinion on anything that I don’t know the other side’s argument better than they do.” — Charlie Munger